This Komen-funded study recruited women with young-onset breast cancer, an unaffected sister already being studied as part of the Sister Study cohort, and, when available, their parents. We will combine their data with the DNA and environmental data now being collected from their unaffected sisters (who previously joined the Sister Study) and saliva-based DNA collected from their parents. We are using a nuclear-family-based approach to study genetic and environmental factors involved in young-onset breast cancer. The study gained enormous operational efficiency advantages, by taking advantage of the infrastructure that was already in place and functioning smoothly for the Sister Study (Dale Sandler, PI). We are almost done with collecting clinical data and validating the diagnoses for all these young-onset cases. Follow-up of these cases (after merging with new cases in the Sister Study) will also allow us to identify environmental, clinical, and genetic factors that influence health after treatment. Case-parent analyses of gene variants are protected against bias due to confounding by genetic heritage, and also permit detection of both maternally-mediated genetic effects and parent-of-origin (imprinting) effects. In the proposed study, the participating affected sisters are each completing a computer-assisted telephone interview like the one their sister completed for the Sister Study, providing information about personal exposures, reproductive history, and past occupational exposures. Environmental effects will be identifiable through a paired comparison of affected and unaffected sisters. Gene-by-exposure interactions will be assessed with novel statistical methods. In summary, the Two Sister Study leverages off the ongoing Sister Study to build a cost-effective, powerful, and statistically independent study of young-onset breast cancer. Findings related to combined effects of genetic variants and environmental factors can be replicated later in the Sister Study. We have completed study enrollment. With augmentation by including some newly diagnosed young-onset cases from the Sister Study we have enrolled nearly 1500 cases providing both questionnaire data and DNA. We have also enrolled 1403 of their parents, who provided DNA. This work was accomplished with assistance from the EB support services contract. We secured permission from the funding agency (Susan G. Komen for the Cure) for a no-cost extension and for redirecting the money originally intended for a candidate gene approach to instead carry out a genome-wide association study (GWAS), using the Illumina OmniExpress plus Exome chip. Carried out through a contract with the Center for Inherited Disease Research, this genotyping project has (augmented by imputations) generated more than 20 million SNPs on these families. We are also participating in the GAME-ON consortium, and this effort has provided additional genotype data based on the newly developed oncochip, again through CIDR. We are using these data to find gene-by-environment causal factors for young onset breast cancer. By combining the Two Sister cases with those incident cases arising in the Sister Study we will also be able to study complexes of factors that are related to healthy recurrence-free survival following treatment. There is interest in a possible relation between history of migraine headache and risk of breast cancer, and we carried out analyses related to that question using data from the Two Sister Study. We distinguished between migraines that tended to occur at a particular time of the menstrual cycle and those that did not follow a menstrual pattern. Overall there was no relationship between migraine and risk of young-onset breast cancer. That paper is now undergoing review. In work with my new post-doc, Katie O'Brien, we considered history of exposure to hormone replacement therapy (HRT) and risk of young onset breast cancer, using data from the Two Sister Study. A complication to overcome in carrying out that analysis was a birth cohort effect related to the publication in 2002 of findings from the Women's Health Initiative trial where the combination estrogen/progesterone hormone replacement therapy was found to increase risk of breast cancer and other conditions. As that report had an effect on clinical practice, women who and gone through menopause prior to 2002 were more likely to have used HRT than were their younger sisters who went through menopause after those findings had been published. Because the control sisters tended to be older than their case sister in the Two Sister Study, this birth cohort effect implied that controls were more likely to have been prescribed hormone replacement therapy. To adjust for this differential opportunity for exposure we carried out a propensity score analysis where we used data from the Sister Study to model the likelihood of exposure and applied those propensity probabilities to the logistic model using data from the Two Sister Study. The resulting paper is now undergoing review. A second paper undergoing review compares the risk factor profiles for ductal carcinoma in situ and invasive breast cancer for women with diagnosis under age 50, again using data from the Two Sister Study. This work was joint with our undergraduate summer intern from last summer, Jenny Sun. In work this past summer with an intern, Denis Whelan, we considered self-reported history of a young-onset eating disorder in relation to risk factors for breast cancer and described associations with later body habits and reproductive events. Analyses of the genetic data are ongoing. We applied for and were awarded a grant from the Office of Dietary Supplements (NIH) to study the role of vitamin D and methylation in breast cancer risk. This work is beginning now, with specimens being pulled for assay.